眼科 ›› 2024, Vol. 33 ›› Issue (4): 241-244.doi: 10.13281/j.cnki.issn.1004-4469.2024.04.001

• 述评 •    下一篇

ChatGPT在眼科的应用与挑战

胡健萍  李建军   

  1. 首都医科大学附属北京同仁医院  北京同仁眼科中心  北京市眼科研究所  眼科学与视觉科学北京市重点实验室100730
  • 收稿日期:2024-06-30 出版日期:2024-07-25 发布日期:2024-07-18
  • 通讯作者: 李建军,Email: jianjunli2005@163.com

Applications and challenges of ChatGPT in ophthalmology

Hu Jianping, Li Jianjun   

  1. Beijing Institute of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University; Beijing Key Laboratory of Ophthalmology and Visual Science, Beijing 100730, China
  • Received:2024-06-30 Online:2024-07-25 Published:2024-07-18
  • Contact: Li Jianjun, Email:jianjunli2005@163.com

摘要: ChatGPT作为预训练生成的大型语言模型正在改变传统的医疗实践,有望显著提高医疗效率与质量。ChatGPT在眼科教育、疾病诊断、病历书写中可以充当眼科医师的辅助工具,也能为眼病患者提供全面、详细且符合临床指南的实用指导。但其在人工智能幻觉、数据安全、隐私保护、模型开发透明性与伦理方面存在潜在的局限性和风险。使用中仍然需要医师的人为监督。眼科医师需了解ChatGPT的性能、水平以及如何利用该技术的独特功能,并能识别其潜在的风险。(眼科,2024,33: 241-244)

关键词:  , ChatGPT;大语言模型;眼科;人工智能

Abstract:  ChatGPT, a pre-trained generative large language model, is set to revolutionize traditional medical practices by significantly enhancing efficiency and quality. In the field of ophthalmology, ChatGPT functions as an auxiliary tool for education, diagnosis, and medical record-keeping, additionally providing comprehensive, detailed, and guideline-compliant practical guidance for patients with eye diseases. However, potential limitations and risks associated with AI hallucinations, data security, privacy protection, transparency in model development, and ethical concerns persist. These necessitate human supervision by physicians during its use. Therefore, ophthalmologists must understand the performance and capabilities of ChatGPT, adeptly leverage its unique features, and remain vigilant to its potential risks. (Ophthalmol CHN, 2024, 33: 241-244)

Key words: ChatGPT, large language model, ophthalmology, artificial intelligence